Machine-learned Fokker-Planck model of nonthermal particle acceleration driven by the kink instability in jets

Recent developments in astrophysical jet simulations suggest magnetohydrodynamic kink instabilities as a mechanism to accelerate particles to nonthermal energies. Though, the underlying processes which give rise to the resulting energy spectrum of the nonthermal particles remains to be fully explored. In this work, we continue exploring a Fokker-Planck-type model of nonthermal particle acceleration powered by the magnetic energy lost in kink instabilities in jets. We utilize a differentiable Fokker-Planck simulation approach to quantify the time-dependence and energy phase-space dependence of advection and diffusion processes and particle escape rate from the jets. Previous work has shown the non-uniqueness of advection and diffusion processes given only a single phase-space distribution of simulation test particles. Current work shows that using multiple particle distributions instead can uniquely resolve said distributions. These results will further elucidate how processes governing nonthermal particle acceleration arise from jet structure and dynamics.